Advanced Rotator Cuff Tear Score (ARoCuS): a multi-scaled tool for the classification and description of rotator cuff tears.

Abstract

To introduce a (semi-)quantitative surgical score for the classification of rotator cuff tears. A total of 146 consecutive patients underwent rotator cuff repair and were assessed using the previously defined Advanced Rotator Cuff Tear Score (ARoCuS) criteria: muscle tendon, size, tissue quality, pattern as well as mobilization of the tear. The data set was split into a training (125 patients) and a testing set (21 patients). The training data set fitted a nonlinear predictive model of the tear score based on the ARoCuS criteria, while the testing data served as control. Based on the scoring results, rotator cuff tears were assigned to one of four categories (ΔV I-IV) and received a stage-adapted treatment. For statistical analysis, mean values ± standard deviation, interclass correlation coefficients (ICC) and kappa values were calculated. Overall, 32 patients were classified as ΔV I, 68 as ΔV II and 37 as ΔV III. Nine patients showed ΔV IV tears. Patients of all ΔV groups improved significantly their Constant scores (p<0.001) and profited from significant pain reduction after surgery (p<0.001). To date, ten patients have undergone revision surgery with five of them primarily classified as ΔV IV. Kappa values for the interobserver reliability ranged between 0.69 and 0.95. ICC scores for the ΔV category were 0.95 for interobserver reliability. The ARoCuS facilitates intra-operative decision-making and enables surgeons and researches to document rotator cuff tears in a standardized and reproducible manner.